Wikipedia talk:WikiProject Astronomy/Archive 37

Imperial/U.S. customary units in the infobox
Should imperial/U.S. customary units be present in the infobox of astronomical objects in general? CactiStaccingCrane (talk) 04:40, 20 January 2024 (UTC)
 * Option 1: No for all articles
 * Option 2: Yes for info cited to sources that only use imperial units, otherwise no
 * Option 3: Yes for all articles


 * Context: Right now, the use of imperial/U.S. customary units are inconsistent among articles (for example, Sun, Mars, Earth and has imperial conversions, but Pluto, Mercury (planet) and Ceres (dwarf planet) don't). For articles that do use imperial/U.S. customary units, they also have SI conversions and often uses convert template.


 * CactiStaccingCrane (talk) 04:45, 20 January 2024 (UTC)
 * MOS:UNITS applies, with the alternative being units conventional to astronomy. Praemonitus (talk) 06:07, 20 January 2024 (UTC)
 * If so then all U.S. customary units be removed from these articles. CactiStaccingCrane (talk) 06:21, 20 January 2024 (UTC)


 * Specifically for science articles (not necessarily for biographies or histories), per MOS:CONVERSIONS: "...in science-related articles, supplying such conversion is not required unless there is some special reason to do so." To avoid conflict, I'd include the linked policy in the edit notes. Also, I'd hesitate to apply this to the Earth article. Praemonitus (talk) 15:14, 20 January 2024 (UTC)

I'm opposed to imperial in general, but we need to keep it when citing values that are given in imperial. The reason is that when we convert to metric, not only may there be rounding errors, but we often change the number of significant digits. And when sources are in imperial, their source data was often in metric and there are already conversion errors involved. Often when we convert back, our figures differ from the original -- that's been a recurring problem with our data. Better to give it in imperial with our metric conversion following in parentheses. Editors will then be aware of the potential for error and try to find the original figures, which should be used instead. When our sources use metric, then we should use metric only, unless our source converted from imperial. In all cases, I think we should attempt to use the original figures, or as close to them as we can find. — kwami (talk) 23:52, 20 January 2024 (UTC)

We should be using metric units for everything except for material which is specifically US material. Wikipdia is not owned by the US. It is world wide and vast majority of countries now use metric. Bduke (talk) 01:11, 21 January 2024 (UTC)


 * Doesn't matter if it's US/NASA. The issue is fidelity to the data. If a US source publishes in metric, we should use metric. If a UK source publishes in imperial, so should we. — kwami (talk) 02:27, 21 January 2024 (UTC)
 * No. We are not writing ransom notes with words cut out of other publications. We are giving coherent explanations that are based on reliable sources. To make our articles coherent, we should choose units appropriate to the article, and give the most appropriate unit first. If the unit given in a source is different from what we choose to list first, we can use the convert template in a way that the value copied from the source listed first and the converted value is given first. Since this thread is about infoboxes, it isn't even necessary to give the value from the source in the box at all, so long as it is in the body of the article. Jc3s5h (talk) 02:35, 21 January 2024 (UTC)
 * That's what I said, at least for the original data appearing first. Unless we explain that the figure given second is primary, and not just added because someone wanted to plaster imperial all over WP. We don't want someone coming by and deleting the data because that's all they think it is. — kwami (talk) 02:39, 21 January 2024 (UTC)
 * convert has a function of just displaying the output. So it is possible to conserve the original data in some way. CactiStaccingCrane (talk) 06:05, 21 January 2024 (UTC)


 * Astronomy data is often published with excessive numbers of digits; far more than is justified by the margin of error. I don't think we need to worry about the accuracy of the conversions. The appearance of excessive accuracy can be misleading in and of itself. We're not an original source for this data, and often we can get by with presenting rounded values. Praemonitus (talk) 05:21, 21 January 2024 (UTC)
 * I agree. That's very often the case, and in such cases conversion errors are not an issue. If we have something at ridiculous (and spurious) precision in imperial, then I have no problem simply converting to metric and ignoring the published units. But not infrequently, especially with initial news reports and even beyond that with crude estimates, data is published to very few sigfigs in imperial, and we do introduce a significant error when converting. Say, the impact of an asteroid est. to be 10 miles in diameter -- what do we convert that to? 15 km? 20 km? Usually I see a misleading 16 km, a precision that is not justified by the source. — kwami (talk) 07:07, 21 January 2024 (UTC)
 * Again, convert can be helpful here, by using "round=5" or "round=25" to round to the nearest .5 or .25. CactiStaccingCrane (talk) 07:11, 21 January 2024 (UTC)
 * Indeed. I just think that in such cases we need to be clear to the reader that the source data is in imperial, not that we just decided to add a conversion to imperial. — kwami (talk) 08:26, 21 January 2024 (UTC)


 * Option 2 seems the right solution to me. If the referenced source is in miles, we should give that value in the infobox for verification, along with a km conversion. If the source was in km, there's no need for miles. A conversion into au or pc might be more useful anyway, depending on context. The sentence about 'science-related articles' in MOS:CONVERSIONS applies and makes sense to me. Modest Genius talk 11:30, 23 January 2024 (UTC)
 * There's a problem though with the wording of Option 2: "Yes for info cited to sources that only use imperial units, otherwise no."
 * The problem is that sources often include a conversion to metric with spurious precision (e.g. 10mi/16km, where the 10mi is the original number and only an estimate), and the wording of Option 2 means we'd cite only that conversion. That conflicts with the spirit of the MOS warning "Be careful especially when your source has already converted from the units you're now converting back to." That's not just a potential problem with converting back, but with dropping the original number and citing only the converted value. In my example, we shouldn't cite "16 km", but rather something more like "10 mi (10-20 km)" or "10 mi (approx. 15 km)", or even just "10-20 km" or "approx. 15 km". — kwami (talk) 12:11, 23 January 2024 (UTC)
 * In those cases, I think it's fine to use the US units in convert with disp=out, so the original is displayed to editors for verification and the metric is displayed to readers for consistency. So yeah, the wording in the MOS should advise using the original quantity and doing our own conversion. -- Beland (talk) 01:54, 18 May 2024 (UTC)
 * Okay, added to the MOS. — kwami (talk) 02:15, 18 May 2024 (UTC)
 * Perfect. I think that addresses all my concerns. (Sorry, I didn't see the extraneous changes that I typed somehow.) — kwami (talk) 02:36, 18 May 2024 (UTC)

RFC: Imperial/U.S. customary units in astronomy object infobox
As a month passes by and Wikipedia_talk:WikiProject_Astronomy did not reach to a consensus, I think it is time to ask the wider community.

CactiStaccingCrane (talk) 16:22, 27 February 2024 (UTC)
 * Option 1: Omit all of imperial/U.S. customary units in infobox, but allow using cvt to convert numbers from the source to SI units
 * Option 2: Only use imperial/U.S. customary units in infobox if the source itself only use imperial/U.S. customary units
 * Option 3: Allow imperial/U.S. customary units in infobox


 * , this is exactly the same question - if there is no firm consensus from one RFC, why do you think a second one will garner it? Primefac (talk) 16:38, 27 February 2024 (UTC)
 * That's because my first RFC is somewhat vague and did not gather enough activity. I do believe that the second RFC will come to a definite consensus. CactiStaccingCrane (talk) 16:42, 27 February 2024 (UTC)
 * None of the above. Do not use imperial/US customary units so that they appear in infoboxes. If the source only provides an imperial/US customary unit, or the imperial/US customary unit is better because the SI unit contains unjustified excess significant figures, include the imperial/US customary unit in the body of the article or in a footnote. Jc3s5h (talk) 16:39, 27 February 2024 (UTC)
 * Can you explain further what you meant by "excess significant figures"? Can we just trim them out for the sake of brevity? CactiStaccingCrane (talk) 16:42, 27 February 2024 (UTC)
 * Sometimes a source will give a value in imperial, such as 93 million miles, which has a few significant figures. The source will give a sloppily-converted value in SI, such as 149.7 Gm in parenthesis. If this value were given in the Wikipedia article it would likely be wrong, because the conversion was calculated from the rounded value, 93 million, rather than the true value. For example if we were talking about the semi-major axis of the Earth's orbit the true value is 92955902 and the value in km, rounded to four digits, would be 149.6 Gm. Jc3s5h (talk) 17:07, 27 February 2024 (UTC)
 * This is similar but not identical to the question you asked a month ago, which is confusing. Option 1 is now closest to my preference, but not quite what I would have chosen. More broadly, where are you proposing this rule would go? With what wording? Is there a problem that needs fixing?
 * In my view, astronomy infoboxes should use a) professional astronomy units (parsecs, solar masses etc.) and b) an SI equivalent if relevant. We don't need to convert everything e.g. listing every star's mass in kilograms isn't helpful to anyone. I wouldn't include an imperial conversion at all, unless the only units available in the references were imperial and the conversion to more appropriate units is shown. I highly doubt that there are many sources that are reliable, only use imperial, and have no alternative better source that gives the value in astronomical or SI units, so that's a very rare edge case. Modest Genius talk 17:49, 27 February 2024 (UTC)


 * As a rule, astronomy infoboxes should contain standard astronomy units, plus SI if relevant (and only if relevant), but no imperial conversion. I'd say that Imperial conversions could be provided in prose, but should never be primary, and shouldn't go in at all unless the number is of significant interest to non-scientists.  I'd add that any rule relying on the specific source adopted - such as Option 2 - is a nonsense because it opens the gate to people just switching the source to something that matches their unit preference.  (And yes, people really do do this.) Kahastok talk 18:24, 27 February 2024 (UTC)
 * We might use only SI in the info box, but imperial in the text if that's what the original (or oldest attested) measures are given in. My worry is the errors that creep in by converting a rough estimate (e.g. 10 miles as a guestimate) into something that looks spuriously precise (16 km). — kwami (talk) 23:47, 27 February 2024 (UTC)
 * As for cherry-picking sources, we want the original data, not imperial conversions. Only if the original data is in imperial should we use imperial. Sometimes the oldest source we have has converted SI to imperial, and later SI might be back converted from the imperial. We don't want to copy that. This is mostly a problem with press releases. But once we can access the original data, we should of course go with that. — kwami (talk) 23:49, 27 February 2024 (UTC)
 * This is all already considered in WP:UNITS, and I'm not sure why this topic should be an exception to the standard rules there? Of course if there is an actual quotation (direct or indirect) then you preserve the quote.  But if the first person who measured the distance to Mars did it in smoots, that would not be a good reason for us to use smoots for the purpose. Kahastok talk 18:21, 28 February 2024 (UTC)
 * If that's the only measure we have, then yes, we should give the results in smoots. I'm not talking about the oldest measurement, but about the original units of the measurement we use. — kwami (talk) 18:40, 28 February 2024 (UTC)
 * My view on each of these proposals:
 * Option 1: Easily the best as there is realistically no context in astronomy in which imperial/USC is preferred over SI (or other scientific units, including occasionally some pre-SI metric relics). My main objection is to the use of the term "source" since, every time a proposal has arisen for source-based units, it has quickly become apparent how problematic and unworkable it would be. It's perhaps best to clarify that (as I understand) what is being suggested here is the use of the convert template with the disp=output only flag, so if a number is quoted in imperial/USC we can still take that information from the source but display it in appropriate units only e.g. 120 mi &rightarrow; 120 mi.
 * Option 2: This seems to be just a more explicit proposal for source-based units, which I would oppose as described above.
 * Option 3: No, as in the real world there's no case in which the very marginal benefit of such unnecessary conventions is worth the clutter added by their inclusion. Our overarching guidance is to provide conversions only where they are, in context, likely to be useful to readers. If someone is hell-bent on knowing the diameters of the moons of Mars in furlongs, they have access to other conversion tools. This also makes me wonder exactly what "allow" means – allow individual editors to do whatever they want, regardless of MOS-level guidance or stylistic consistency across astronomy articles?
 * So if we were to adopt any of these, my !vote would be for a clarified version of the first. A more general comment would be that, as others have observed above, this is spelled out in realistically enough detail already at WP:UNITS, so any guidance specific to this WikiProject should simply be a clarification of what that MOS guidance means specifically for the articles it maintains. It's important not to focus obsessively on problems that don't exist, or don't really manifest in article-space, so unless there is a recurrent problem with editors obsessively adding imperial/USC units to astronomy infoboxes, I'm not sure see a clear use case for this guidance. Archon 2488 (talk) 15:22, 1 March 2024 (UTC)
 * Option 1. I favor being slightly stricter than MOS:UNITS and discouraging US/imperial units in infoboxes about astronomical objects, rather than just not requiring converting to them. Kwamikagami has just added (and I tweaked) advice to MOS:CONVERSIONS which should address the concerns raised above about rounding and significant figures and retaining the original source units for verification (e.g. use disp=out for US/imperial quantities we want to keep for editor verification but display to readers in metric only). This means there is no reason to use US/imperial just because a source does so. Infoboxes are crowded, and the quantities involved are usually mind-blowingly large anyway, so it's more important to convert to the astronomical system of units than to US/imperial units. So perhaps some wording like this:
 * "In infoboxes for astronomical objects, follow MOS:CONVERSIONS making the International System of Units and astronomical system of units primary, but do not display US conventional/British imperial units. Display conversions from SI to astronomy units if appropriate."
 * -- Beland (talk) 03:15, 18 May 2024 (UTC)
 * I agree about restricting Imperial. What does it mean to anyone to say that Jupiter "weighs" X pounds? (Okay, we have pound mass now, but still.) No-one has any comprehension of what numbers that large mean. They're only useful for comparison, and units don't matter for that. The only figure likely to mean anything to anyone is that the Earth is 93M miles from the Sun, which I learned as a child, but the AU takes care of that: say that Jupiter is at 5 AU, and is doesn't matter if someone memorized the AU in miles or km. Lots of people have memorized the circumference of the Earth is 24k miles, but we hardly mention circumferences in the info boxes. Other than that, imperial is pretty meaningless.
 * I'd like to take it further, though. People have been going around converting exponential notation to SI prefixes (terameters and zetagrams or whatever). Those are also pretty much useless. IMO we should use exponential notation with basic SI (MKS) units.
 * Personally, I prefer the engineering convention of restricting exponents to multiples of three, even though my background is in physics, because it makes comparisons easier to have the same scale for similar objects. Certainly in tables (e.g. most massive objects in the SS), the header should specify the exponent so that the columns can be sorted easily, but I think it would be useful even in info boxes. But even if we go by the physicist convention, that's better than SI prefixes. — kwami (talk) 03:35, 18 May 2024 (UTC)

Wikipedia:Manual of Style/Astronomy
I put together an early draft of a MoS guide for astronomy under WikiProject Astronomy/Manual of Style. It is intended to embed what has thus far been tribal knowledge for this WikiProject and its associated task forces. For comparison, other such style guides can be viewed under the "By topic area" tab in the infobox. What do you think of this proposed guide? Do you disagree with what is stated? What else should it cover? I'm sure it can be significantly expanded. Thanks. Praemonitus (talk) 05:24, 11 April 2024 (UTC)
 * This is very helpful. I particularly appreciate the linking to other relevant guidelines. ~ Maplestrip/Mable ( chat ) 06:40, 11 April 2024 (UTC)
 * I would suggest adding something discouraging the usage of computer-generated 2d/3d models in the lead image when suitable alternatives are available; this prominently applied to Miranda, where a 3d model was used as the infobox image for several years in lieu of an image or image mosaic.
 * It may also be worth mentioning an apparent informal convention to use full-disc images of visited Solar System objects when possible, as the Sun, all nine planets (except for Venus, which does not have any full-disc true-color images on commons), Pluto, Ceres, and all visited moons seem to follow this convention. A similar/analagous convention seems to be in place for imaged galaxies and DSOs too. ArkHyena (talk) 00:14, 15 April 2024 (UTC)
 * Makes sense. I've seen truncated images from Hubble in the lead image spot, which might not be ideal. A pair of examples are NGC 5506 and NGC 5634. Praemonitus (talk) 04:45, 15 April 2024 (UTC)
 * Of course! There may also be an additional caveat needed in that quality/accuracy still has priority over full disc. To provide a planetary example, this quarter-full image of Europa takes precedence over this full disc image, which is lower-resolution and exaggerated color. ArkHyena (talk) 20:58, 16 April 2024 (UTC)


 * As there didn't appear to be any serious objection to this proposal, I added it to the project sidebar so it can cook longer. Eventually, I hope it can be proposed as an addition to the MoS. Praemonitus (talk) 18:22, 30 April 2024 (UTC)

Abbreviations
A style question came up during a recent edit discussion: should infobox data entries preferentially use abbreviations or words? For example, 'Mly' or 'million light years'; 'AU' or 'astronomical unit'; 'Gyr' or 'billion years'; 'g/cm3' or 'grams per cubic centimetre'. In my mind the infobox should be kept compact by sticking to abbreviations, with the word usage being left for the article body. Is there a preference? Praemonitus (talk) 20:42, 12 April 2024 (UTC)
 * Okay, it's already covered by the MoS: "Where space is limited (such as tables, infoboxes, parenthetical notes, and mathematical formulas) unit symbols are preferred." Praemonitus (talk) 22:17, 12 April 2024 (UTC)

AI
It's probably not an issue yet, but should we mention AI-generated illustrations? For example, "AI-generated illustrations should be avoided unless their accuracy is confirmed by an astronomy expert. The AI system may have been trained using copyrighted material, so the legality of such use is unclear." Praemonitus (talk) 17:28, 17 April 2024 (UTC)


 * New here so I don't know if it has already been covered in a more general guide or such, but I believe mentioning AI-generated images is a good idea. It's better to cover everything before it becomes a problem, than having to adjust it later in my opinion. AstroChara (talk) 17:15, 18 April 2024 (UTC)


 * The copyright rules for AI-generated images is AI-generated media. If we prepare rules for objects that have never been photographed (such as exoplanets), gone millions of years ago (Theia, Mars with oceans), or just hypothetical (a terraformed Mars), they should be for any way to create such an image, be it AI, an artist's work or whatever. Cambalachero (talk) 19:40, 18 April 2024 (UTC)
 * I guess that artists and the astronomy experts are also trained using copyrighted material. Is the legality of the use of their work also unclear? Johnjbarton (talk) 23:05, 18 April 2024 (UTC)
 * You're taking the piss, right? Primefac (talk) 11:31, 19 April 2024 (UTC)
 * I think you are misunderstanding the issue with source materials and AI. Artists in all genres and authors in all fields re-mix material. That's Standing on the shoulders of giants. The legality of that re-mix is never clear cut. That's why we have courts, law suits, and lawyers. Furthermore, the legal onus is on the person who causes the material to be displayed, not on the creator. I can copy Monet in my home all day and night. Only when I sell my painting on the street will it be illegal. Similarly, the AI generation won't be illegal, uploading the image that infringes on a copyright will be. Johnjbarton (talk) 15:33, 19 April 2024 (UTC)
 * The legality of that is already established via the ramifications of plagiarism and the fair use law. No such ruling exists for AI; it's not even clear they can generate patents. Praemonitus (talk) 14:21, 19 April 2024 (UTC)
 * The discussion was copyright, not patent. Johnjbarton (talk) 15:21, 19 April 2024 (UTC)
 * Thanks for stating the obvious. I was making the point that the legalities on the use of AI output are far from settled. Praemonitus (talk) 19:26, 19 April 2024 (UTC)
 * I do agree here. The morals and merits of AI art can be debated here endlessly, but the primary concern is that laws surrounding AI art (and really, AI anything) are young and volatile, and potentially subject to rapid near-future changes. It would probably be wise to at least require AI-generated material to be clearly marked as such. ArkHyena (talk) 20:54, 19 April 2024 (UTC)


 * As for the way an AI generates an image, unless there is an actual case law over that, I would dismiss the whole thing of "they used copyrighted images for training" as immaterial. The only thing that is truly relevant is the final image, the result of the prompt. If that image is derivative, then delete (or don't upload), under the same conditions we would with any other non-AI image. Cambalachero (talk) 14:12, 19 April 2024 (UTC)
 * My concern is that there is current controversy about the use of copyrighted web content for training, which may ultimately lead to take-down orders. Praemonitus (talk) 12:37, 20 April 2024 (UTC)
 * Praemonitus: I like the way you worded this, and I agree with a fairly broad restriction on use of AI images without prior vetting and/or secondary sources. - Parejkoj (talk) 16:02, 19 April 2024 (UTC)

SIMBAD and NED
The guide contained the following entry, which was removed with this edit:
 * However, in general they should not be cited as a source for the infobox data because the information is subject to change over time and comes from multiple overlapping sources. However, they are useful as a stable reference for the other designations listing.

with the comment "I disagree, it is better to directly cite SIMBAD or NED as sources for the infobox, as it makes it easier to verify the information."

I have seen cited data removed from these sources, so they should not be considered stable. An example of this is the coordinate information, which is subject to refinement over time, with the old data being replaced. Instead, in many cases they do provide stable references that can be used to directly cite the data. Hence, I'd caution against using SIMBAD or NED directly.

Are there any concerns about this? Praemonitus (talk) 12:35, 20 April 2024 (UTC)

Units
A few suggestions: -- Beland (talk) 03:51, 18 May 2024 (UTC)
 * I think it would be good to pick one symbol for astronomical unit, and I think right now we almost always use "AU".
 * Surface gravity should probably be displayed in standard gravity units in addition to m/s2, and it would be nice to use "g0" instead of "g" to avoid confusion with grams.
 * Conversions between multiple SI expressions should be discouraged (e.g. both km/h and m/s, which I do see).
 * This section is a bit unclear on whether both SI and astronomy units should be used, or only one. In infoboxes (and prose where it's not excessive) I would argue for both. It's much easier to intuitively grasp huge distances and masses in AU, light-years, Earth gravity, Sun masses, etc., so very worthwhile to convert into these if necessary. I like the idea of converting to metric units as well because it allows intuitive comparisons across measurements - for example, how much bigger is the Earth than its orbit around the Sun? And as much as I want everyone to learn the scientific units from birth, it's a bit kinder to Americans who have come halfway and learned either metric or astronomy units but still haven't gotten a handle on the other.
 * I'm agnostic as to whether metric or astronomy units should be primary where both are given via conversion, and I'm actually fine with just making whichever the cited source is using as primary for convenience. But whatever the rule is it should be made clear up front.